Study Reveals Why You Got those Ads

We’ve all been served ads that made us wonder, “How’d that sucker get in my feed?” The answer is inefficiency. Adlook, a global media tech company specializing in digital advertising solutions, conducted a study that showed how outdated data results in reaching the wrong audiences with ads.

The chief offender is demographic targeting.  In a poll of more than 1,300 people in the U.S., respondents indicated that they had received ads well outside their socio-demographic. For example, renters received digital ads clearly intended for homeowners. Ads intended for women aged 18-24 reached 18% of that intended audience, while 43% of those ads went to men, and 35% went to audiences aged 55. Seventy percent of people targeted as parents, had no children.

Mateusz Jedrocha, Chief Product Officer at Adlook, said brands should rethink how they define and reach audiences by moving from demographic-based targeting to behavior-based models. They could also prioritize contextual intelligence over data segments and test AI-driven audience segmentation to identify real-time consumer interests.

Is there “harm” in targeting non-parents with content meant for people who have kids? Many childless folks at least have children in their lives in the form of nieces, nephews, and friends’ kids.

It’s not necessarily about “harm,” but about efficiency and effectiveness in advertising. While some non-parents may have children in their lives, brands investing in targeted advertising expect their messaging to reach actual parents – not just individuals with a secondary connection to children. The core issue is that advertisers are paying a premium for inaccurate data that fails to deliver on its promise. Misalignment leads to wasted ad spend, lower engagement rates, and ultimately, reduced campaign performance.

Ditto for targeting “homeowners” who are actually renters. Maybe they aspire to own a home one day.

It’s rarely binary. We at Adlook can absolutely imagine situations where the difference between owner and renter is irrelevant (e.g., “target people with a long commute”). Yet, aspiring to own a home and actually being a homeowner are two very different consumer states. If a brand is marketing mortgages or home-improvement services, reaching actual homeowners is crucial for effectiveness. The problem is that 40% of individuals categorized as homeowners in our study were actually renters – a significant misclassification that leads to misplaced marketing dollars.

How can brands target audiences based on their interests, which the study seems to suggest is more effective?

Planning impressions based on context is not a new concept. What is new is emerging technologies, which allow for much better contextual placements at scale. Large language models (LLM) & Natural-Language Processing (NLP) unlock two great possibilities: A much better understanding of marketing audiences, which are never as simple as “women 18-34,” but rather rich descriptions written in natural language. Also, a far superior understanding of website content and, thus, for whom this content may be relevant.

For example, an article titled “Top 5 Family SUVs” on an automotive site may have been classified as “automotive context” by a simple keyword-based context engine. However, an LLM/NLP engine will understand that this article is highly relevant for any family who likes spending time together, enjoys family trips, and perhaps engages in active forms of recreation.

Advanced contextual engines open entirely new inventory for many advertisers—inventory that, until now, was largely “invisible” to them, as it was classified in a simplistic, “flat” manner.

To what extent does search marketing solve for these issues?

Search marketing is a very broad term. When talking about contextual advertising, Adlook refers to a process in which we analyze what websites answer a specific user search query and then seek to advertise on those sites. In other words, we’re still talking about display or video advertising—not SEM/SEO. Search marketing is more intent-driven, making it a stronger alternative to socio-demographic targeting. When a consumer actively searches for “best strollers for newborns,” there’s an immediate, self-declared intent—something demographic labels can’t match.

However, search marketing has its limitations, including limited scale and competitive costs that drive up CPCs. To build a well-rounded strategy, brands should combine search marketing with contextual targeting and AI-driven insights, enabling them to reach both active searchers and passive yet relevant audiences effectively.

What about OOH? Would brands do better to serve messages in this location-based context?

It’s interesting how contextual advertising in programmatic has high resemblance to how we plan other, often “traditional” media. Out-of-Home is a perfect example. OOH advertising has several advantages over flawed socio-demographic targeting in programmatic:

It provides real-world relevance. OOH placements are location-based, making them contextually relevant (e.g., placing a baby product ad near a daycare). There’s no reliance on personal data, and unlike digital ads, OOH does not depend on cookies or third-party data, making it a privacy-first alternative. Broad reach and brand awareness are great for high-impact storytelling, but with ability to “close the loop” via QR codes, NFC, or other digital extensions.

However, OOH (as many “traditional” media) has some limitations:

• Limited personalization. Unlike digital, OOH does not allow precise targeting based on real-time intent. It is now changing with Digital OOH to some extent though!
• Higher costs for premium placements. Prime locations can be expensive, especially in major cities.

A best practice is a hybrid approach, combining OOH for broad awareness and contextual/dynamic digital strategies for engagement and conversion.